Survival outcomes of patients after liver transplantation for patients with hepatocellular carcinoma exceeding the Hangzhou criteria Authors: Zhi-Ping Huang, Shu-Tong Lu, Jin Du, Ren-Dong Liu, Qing Ouyang, Bao Zhang, Meng-Chao Wang, Diamantis I. Tsilimigras, Hani M. Wadei, Yu-Jian Zheng, Shao-Ping Wang, Feng Huo, Xiao-Yu Tan, Jian-Xiong Chen Published: 2025 ABSTRACT BackgroundPatient selection is one of the key factors influencing prognosis after liver transplantation (LT), and a number of selection criteria have been proposed and broadly applied. The benefit of LT is debatable when patients exceed these criteria in light of liver worldwide donor shortage. This study aims to identify key prognostic indicators predicting patients undergoing LT beyond Hanzhou criteria, to guide LT candidate selection.MethodsThe clinical data and pathological findings of patients who underwent LT in the Liver Transplantation Center of General Hospital of Southern Theater Command from September 2003 to August 2017 were collected for retrospective analysis, only patients exceeding the Hangzhou criteria were enrolled. Patients were followed up until December 1, 2023; disease-free survival (DFS) and overall survival (OS) was analyzed; and predictive models of DFS and OS were constructed and validated.ResultsA total of 161 patients were included in the analysis. A tumor diameter ≥6 cm [hazard ratio (HR) =2.01; P=0.01], alpha fetoprotein (AFP) levels >1,000 ng/mL (HR =2.68; P<0.001), lack of metformin administration (HR =2.98; P=0.008), operative blood loss >2,000 mL (HR =1.75; P=0.01), and female donor gender (HR =3.71; P=0.004) were independent risk factors for lower DFS, while a tumor diameter ≥6 cm (HR =2.05; P=0.001), AFP level >1,000 ng/mL (HR =1.54; P=0.04), and female donor gender (HR =4.45; P<0.001) were independent predictors of worse OS. Patients with a preoperative AFP level ≤1,000 ng/mL or a history of metformin administration had a significantly better prognosis after LT.ConclusionsFor patients exceeding the Hangzhou criteria, those with an AFP level ≤1,000 ng/mL can still achieve acceptable long-term prognosis after LT. Administration of metformin has a strong positive association with longer recurrence-free survival. The patient criteria for selection for LT can be tentatively expanded. Highlight box Key findings • For patients exceeding the Hangzhou liver transplantation (LT) criteria, those with an alpha fetoprotein level ≤1,000 ng/mL can achieve satisfying long-term prognosis after LT. • Administration of metformin may lower the recurrence rate of liver cancer after LT. What is known and what is new? • Patients undergoing LT for hepatocellular carcinoma within the criteria have a favorable prognosis. • The criteria of patient selection for LT can be tentatively expanded. What is the implication, and what should change now? • Administration of metformin for patients undergoing LT has a strong positive connection with better recurrence-free prognosis, the potential mechanism of which should be explored by more well-designed investigations. Introduction Hepatocellular carcinoma (HCC) has been a persistent threat to human health, with a high cancer-related mortality imposing a considerable burden on medical resources (1). A number of risk factors related to HCC have been reported, such as hepatitis B and/or C infection, metabolic associated fatty liver disease, alcohol, and aflatoxins (2). However, most patients with HCC are diagnosed at a late stage of the disease, thus limiting the available therapeutic options and their related efficacy, with the median survival among patients with untreated HCC being less than 1 year (3). Liver transplantation (LT) is an effective treatment option for the treatment of end-stage liver disease, and appropriate patient selection is an important factor affecting the prognosis of LT recipients (4-8). In order to obtain a better therapeutic effect and make rational use of donor liver resources, various LT centers have put forward different selection criteria for patients with HCC. Among these, the Hangzhou criteria were first to resolve the limitation of the tumor morphological index and are based on China’s national conditions and a combination of pathological and biological characteristics (9). The Hangzhou criteria consist of three elements: (I) the sum of tumor diameters ≤8 cm, (II) the sum of tumor diameters >8 cm but with alpha fetoprotein (AFP) ≤400 ng/mL, and a moderately or well-differentiated grade, and (III) no intrahepatic macrovascular invasion or extrahepatic metastasis (10,11). The number of patients with HCC in China is high, and most are complicated with posthepatitic cirrhosis (12-14). Among these patients with poor underlying liver function, surgical resection is frequently prohibitive, and LT is one of the curative options for HCC. Early clinical experiences in China suggest that some patients with advanced HCC can achieve long-term disease-free survival (DFS) after transplantation (15). However, LT exceeding the criteria for liver cancer still needs to be supported by additional investigations based on a high level of evidence. In this study, the clinical and pathological data of patients with HCC exceeding the Hangzhou criteria who underwent LT in the Liver Transplantation Center of General Hospital of Southern Theater Command from September 2003 to August 2017 were collected to analyze risk factors affecting prognosis and to provide a reference basis for LT candidate selection. We present this article in accordance with the STROBE reporting checklist (available at https://jgo.amegroups.com/article/view/10.21037/jgo-2025-aw-864/rc). Methods Ethics statement All clinical data were obtained with the informed consent of patients and with approval from the Clinical Research Ethics Committee of General Hospital of Southern Theater Command (No. 2024GJJ064). The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. Immunosuppression protocol All patients received a standardized triple-drug immunosuppression regimen starting within 24 hours post-LT, which was formulated in accordance with the diagnosis and treatment specification for immunosuppressive therapy and rejection of LT in China (2019 edition) (https://www.organtranspl.com/article/doi/10.3969/j.issn.1674-7445.2021.01.002). The regimen consisted of tacrolimus, mycophenolate sodium, and short-term corticosteroids. Inclusion and exclusion criteria of patients receiving LT The inclusion criteria for patients in this study were as follows: (I) first LT; (II) primary HCC diagnosed by pathology; (III) exceeding Hangzhou criteria; and (IV) complete clinical data of donors and recipients. Exclusion criteria were: (I) secondary or multiple LT; (II) combined multiple organ transplantation; (III) split LT; (IV) history of other malignant tumors; (V) incomplete clinical data; (VI) complications of portal vein or hepatic vein tumor thrombus; (VII) extrahepatic metastasis; and (VIII) LT of liver cancer according to the Hangzhou criteria. Data collection The clinical data of patients exceeding the Hangzhou criteria who underwent LT in the Liver Transplantation Center of General Hospital of Southern Theater Command from September 2003 to August 2017 were retrospectively collected. Tumor-related indices including tumor number, tumor size, macrovascular invasion, and tumor cell differentiation grading were collected. Microvascular invasion and tumor cell differentiation grading were evaluated in accordance with the pathological findings. The judgment regarding fulfillment of the Hangzhou criteria was based on pathological findings. The candidates’ clinical data included age, gender, and hepatitis B virus (HBV) infection status. History of metformin administration was defined as continuous and standardized usage of metformin for more than 1 year immediately before LT. After LT, all the candidates underwent regular follow-up. Serum AFP was examined and ultrasonography was employed at monthly intervals in the first year and then every 3 months thereafter if no recurrence was observed. Additionally, chest computed tomography (CT) was routinely performed concurrently with either magnetic resonance imaging (MRI) or enhanced abdominal CT scans—these combined imaging assessments (chest CT + abdominal imaging) were scheduled every 3 months in the first 2 years post-transplant, and then every 6 months thereafter. The observation indicators included the following: (I) DFS and overall survival (OS) of all patients for 0.5, 1, 3, and 5 years; (II) univariate and multivariate analysis for risk factors associated with 5-year DFS; (III) univariate and multivariate analysis for risk factors associated 5-year OS; (IV) and a comparison of tumor-free survival and OS among different AFP level groups. Follow-up was conducted in the outpatient clinic or by telephone on a regular basis to record the recurrence, metastasis, and survival of the patients. The patients were followed up once a year until December 1, 2023. Statistical analysis SPSS 24.0 (IBM Corp., Armonk, NY, USA), GraphPad Prism 9.0 (GraphPad Software, La Jolla, CA, USA), and R 4.1.2 (The R Foundation for Statistical Computing) was used in data analysis. Continuous variables are expressed as the mean ± standard deviation. Categorical variables are presented as frequency and proportions and were compared with the χ2 test or Fisher exact test when appropriate. The cutoff value of certain parameters was determined using receiver operating characteristic (ROC) curve analysis. The independent samples t-test or rank sum test was used for comparison between two groups, while the chi-square test was used for the comparison of categorical variables between groups. The Kaplan-Meier method was used to analyze postoperative survival statistics. The Mantel-Cox test was applied to evaluate the difference in survival outcomes. A Cox regression model was used to analyze the independent risk factors associated prognosis, which were used to construct a nomogram that was validated with the concordance index and calibration curve. Results with a P value <0.05 were considered to be statistically significant. Results Clinical and pathological features of patients involved A total of 161 patients were included in this study. In terms of recipients, there were 150 males and 11 females, the mean age was 47.80±11.22 years, and the mean BMI was 22.43±2.94 kg/m2. Regarding donors, 146 cases were male, 15 cases were female, the mean age was 31.77±8.43 years, and the mean BMI was 22.83±3.61 kg/m2 (Table 1). Median follow-up was 2.17 years, with the minimum follow-up period of 9 days and the longest follow-up period was 15 years. Table 1Clinical and pathological features of patientsParameterAll (n=161)Disease-free survivalOverall survivalNon-recurrence (n=53)Recurrence (n=104)PSurvival (n=50)Death (n=105)PRecipient Sex0.74>0.99 Male150 (93.2)49 (92.5)95 (94.1)47 (94.0)98 (93.3) Female11 (6.8)4 (7.5)6 (5.9)3 (6.0)7 (6.7) Age, years47.80±11.2250.79±10.0846.41±11.860.0248.18±10.7247.80±11.750.84 BMI, kg/m222.43±2.9422.83±3.5322.19±2.550.2022.94±3.4722.19±2.630.14 Tumor diameter, cm7.77±5.017.24±5.748.16±4.630.307.56±4.847.95±5.090.67 Number of nodules5.49±5.075.17±5.135.56±5.120.663.98±4.056.08±5.370.01 AFP, ng/mL5,932.61±15,654.523,917.19±11,518.257,375.66±17,794.820.202,111.03±5,507.578,066.74±18,690.380.003 HBV history0.130.27 No135 (83.9)45 (93.8)*86 (85.1)*42 (93.3)*90 (85.7) Yes21 (13.0)3 (6.3)15 (14.9)3 (6.7)15 (14.3) Diabetes0.0050.048 No138 (85.7)37 (77.1)*94 (93.1)*36 (80.0)*96 (91.4) Yes18 (11.2)11 (22.9)7 (6.9)9 (20.0)9 (8.6) Disease history (multiple types)0.130.07 No106 (65.8)37 (90.2)*65 (79.3)*35 (92.1)*68 (79.1)* Yes24 (14.9)4 (9.8)17 (20.7)3 (7.9)18 (20.9) Abdominal surgery history (all types)0.0060.002 No120 (74.5)33 (62.3)83 (82.2)30 (60.0)87 (82.9) Yes41 (25.5)20 (37.7)18 (17.8)*20 (40.0)18 (17.1) Neoadjuvant therapy0.820.02 No65 (40.4)19 (39.6)*42 (41.6)*12 (26.7)*50 (47.6) Yes91 (56.5)29 (60.4)59 (58.4)33 (73.3)55 (52.4) Posttarget drug<0.0010.44 No134 (83.2)53 (100.0)74 (73.3)*43 (86.0)85 (81.0) Nexavar27 (16.8)0 (0.0)27 (26.7)7 (14.0)20 (19.0) Operative blood loss, mL2,791.60±2,286.133,410.42±2,897.562,549.41±1,949.310.072,255.11±1,435.353,063.90±2,572.280.02Donor Sex0.370.18 Male146 (90.7)46 (88.5)*94 (93.1)*47 (95.9)*94 (89.5) Female14 (8.7)6 (11.5)7 (6.9)2 (4.1)11 (10.5) Age, years31.77±8.4332.44±9.2631.65±8.270.6133.49±10.9631.19±7.140.20 BMI, kg/m222.38±3.6121.75±2.6023.00±4.270.1522.80±4.3422.00±2.620.36Data are presented as n (%) or mean ± standard deviation. *, some values were missing. AFP, alpha fetoprotein; BMI, body mass index; HBV, hepatitis B virus. Clinical and pathological features of patients Data are presented as n (%) or mean ± standard deviation. *, some values were missing. AFP, alpha fetoprotein; BMI, body mass index; HBV, hepatitis B virus. OS and DFS At a median follow-up of 2.2 years after LT for HCC exceeding the Hangzhou criteria, 151 patients (94%) had developed recurrence, among whom 144 cases died and 7 cases survived at last follow-up. Median time to tumor recurrence was 6.4 months. There was no recurrence or metastasis in 10 cases, all of whom survived until the end of follow-up. The 1-, 3-, and 5-year DFS of all patients was 31%, 26% and 22%, respectively, while the 1-, 3-, and 5-year OS was 54%, 23% and 17%, respectively (Table 2). Table 2Cumulative survival proportion at different time pointsTime pointDisease-free survival, %Overall survival, %OverallAFP ≤1,000 ng/mLAFP >1,000 ng/mL OverallAFP ≤1,000 ng/mLAFP >1,000 ng/mL 0.5 year4863297677761 year3143155458492 years2837152941163 years2633152333124 years2226152128125 years222615172410AFP, alpha fetoprotein. Cumulative survival proportion at different time points AFP, alpha fetoprotein. Univariate and multivariate analysis of DFS To determine the clinical significance of risk factors associated with DFS and OS, continuous variables underwent bivariate stratification according to cutoff values, which were based on clinical experience and ROC analysis. The cutoff values of AFP levels, age, and operative blood loss were calculated and set as 1,000 ng/mL, 50 years old, and 2,000 mL, respectively. Results of univariate analysis showed that recipient age >50 years old [hazard ratio (HR) =0.62; P=0.04], tumor diameter ≥6 cm (HR =2.33; P=0.001), AFP level >1,000 ng/mL (HR =2.51; P<0.001), and lack of metformin administration (HR =2.43; P=0.03) were significantly associated with lower DFS of the patients after LT. The above parameters, along with the clinically fundamental index of operative blood loss and donor gender, were input into the multivariate Cox regression. The results of multivariate Cox analysis showed that tumor diameter ≥6 cm (HR =2.01; P=0.01), AFP level >1,000 ng/mL (HR =2.68; P<0.001), lack of metformin administration (HR =2.98; P=0.008), operative blood loss >2,000 mL (HR =1.75; P=0.01), and female donor gender (HR =3.71; P=0.004) were independent risk factors for worse DFS after LT (Table 3). Table 3Risk factor analysis of disease-free survival according to the Cox regression modelParameterUnivariateMultivariateHR (95% CI)PHR (95% CI)PRecipient Gender (female vs. male)0.72 (0.26–1.96)0.52 Age >50 years0.62 (0.40–0.97)0.040.74 (0.46–1.18)0.20 BMI, kg/m20.99 (0.91–1.07)0.74 Tumor diameter ≥6 cm2.33 (1.42–3.80)0.0012.01 (1.17–3.44)0.01 Number of nodules1.05 (0.99–1.10)0.09 AFP >1,000 ng/mL2.51 (1.61–3.91)<0.0012.68 (1.65–4.34)<0.001 HBV history (yes vs. no)0.75 (0.30–1.87)0.54 Metformin administration (no vs. yes)2.43 (1.12–5.30)0.032.98 (1.33–6.67)0.008 Abdominal surgery history (yes vs. no)0.59 (0.34–1.04)0.07 Neoadjuvant therapy (yes vs. no)0.73 (0.47–1.14)0.17 Preoperative TACE1.35 (0.84–2.13)0.22 Operative blood loss >2,000 mL1.53 (0.99–2.36)0.061.75 (1.12–2.75)0.01Donor Gender (female vs. male)1.95 (0.84–4.51)0.123.71 (1.51–9.13)0.004 Age, years1.01 (0.99–1.04)0.30 BMI, kg/m21.07 (0.99–1.15)0.07AFP, alpha fetoprotein; BMI, body mass index; CI, confidence interval; HBV, hepatitis B virus; HR, hazard ratio; TACE, transcatheter arterial chemoembolization. Risk factor analysis of disease-free survival according to the Cox regression model AFP, alpha fetoprotein; BMI, body mass index; CI, confidence interval; HBV, hepatitis B virus; HR, hazard ratio; TACE, transcatheter arterial chemoembolization. Univariate and multivariate risk factor analysis of OS The results of univariate analysis showed that tumor diameter ≥6 cm (HR =1.85; P=0.003), AFP level (HR =1.55; P=0.03), and donor gender (HR =3.70; P<0.001) were significantly associated with the OS of the patients. The results of multivariate Cox analysis indicated that a tumor diameter ≥6 cm (HR =2.05; P=0.001), an AFP level >1,000 ng/mL (HR =1.54; P=0.04), and female donor gender (HR =4.45; P<0.001) were independent risk factors for lower OS (Table 4). Table 4Risk factor analysis of overall survival according to the Cox regression modelParameterUnivariateMultivariateHR (95% CI)PHR (95% CI)PRecipient Sex (female vs. male)0.97 (0.45–2.08)0.93 Age >50 years1.02 (0.69–1.50)0.941.17 (0.77–1.78)0.46 BMI, kg/m21.02 (0.94–1.10)0.66 Tumor diameter ≥6 cm1.85 (1.23–2.79)0.0032.05 (1.33–3.16)0.001 Number of nodules1.01 (0.98–1.05)0.45 AFP >1,000 ng/mL1.55 (1.05–2.28)0.031.54 (1.02–2.31)0.04 HBV history (yes vs. no)0.65 (0.38–1.13)0.13 Metformin administration (no vs. yes)1.77 (0.89–3.51)0.101.75 (0.87–3.56)0.12 Abdominal surgery history (yes vs. no)0.63 (0.38–1.05)0.08 Neoadjuvant therapy (yes vs. no)0.74 (0.50–1.08)0.12 Nexavar (yes vs. no)1.02 (0.62–1.66)0.95 Operative blood loss >2,000 mL1.34 (0.91–1.98)0.141.28 (0.86–1.90)0.23Donors Sex (female vs. male)3.70 (1.95–7.03)<0.0014.45 (2.22–8.90)<0.001 Age, years1.00 (0.97–1.03)0.84 BMI, kg/m20.92 (0.80–1.07)0.27AFP, alpha fetoprotein; BMI, body mass index; CI, confidence interval; HBV, hepatitis B virus; HR, hazard ratio. Risk factor analysis of overall survival according to the Cox regression model AFP, alpha fetoprotein; BMI, body mass index; CI, confidence interval; HBV, hepatitis B virus; HR, hazard ratio. Patients with preoperative AFP ≤1,000 ng/mL or history of metformin administration demonstrated better prognosis after LT The prognosis of patients with a preoperative AFP level ≤1,000 ng/mL was significantly better than those with a preoperative AFP level >1,000 ng/mL in terms of both DFS (P<0.001; Figure 1A) and OS (P=0.04; Figure 1B). Importantly, as for individuals with a preoperative AFP level ≤1,000 ng/mL, remarkable improvement in cumulative survival proportion was observed throughout the follow-up. The 1-, 3-, and 5-year DFS of patients with a low AFP level was 43%, 33%, and 26%, respectively, whereas for patients with a high AFP level, the 1-, 3-, and 5-year DFS rates were 15%, 15%, and 15%. The 1-, 3-, and 5-year OS of patients with a low AFP level was 58%, 33%, and 24%, respectively, whereas for patients with a high AFP level, the 1-, 3-, and 5-year OS rates were 49%, 12%, and 10% (Table 2). In further analysis of the correlation between metformin administration and DFS and OS, all patients were stratified into two groups based on history of metformin administration. Significant improvement in DFS was noted among the group of metformin administration (P=0.02; Figure 1C). Although the metformin administration group showed a markedly superior OS, statistical analysis did not indicate this difference to be significant (P=0.25; Figure 1D). This may be due to the limited sample size in our study. Patients with a preoperative AFP level ≤1,000 ng/mL or a history of metformin administration achieved better prognosis after liver transplantation. All patients undergoing liver transplantation exceeding the Hangzhou criteria were included, with the longest follow-up period exceeding 10 years. Significantly better DFS (A) and OS (B) in patients with a serum AFP level ≤1,000 ng/mL were observed. Remarkably better DFS (C) and non-significantly better OS (D) was observed in the metformin administration group. These findings indicated that for patients exceeding Hangzhou criteria, AFP level and history of metformin administration may both be key factors for predicting the degree of malignancy and patient prognosis. AFP, alpha fetoprotein; DFS, disease-free survival; OS, overall survival. Construction and validation of a risk factor model for predicting the prognosis of patients undergoing LT Further investigation of the prediction value of independent risk factors was explored using the ROC curve for DFS (Figure 2A) and OS (Figure 2B). Tumor diameter ≥6 cm, an AFP level >1,000 ng/mL, lack of metformin administration, operative blood loss, donor gender, and a composite index combining the above five predictors were included in the ROC analysis for DFS. The composite index was established using multivariate regression analysis. The highest area under the curve (AUC) was achieved by the composite index (AUC =0.7088), followed by AFP level (AUC =0.5975) and metformin administration (AUC =0.5799). The AUCs of donor gender and operative blood loss were 0.5230 and 0.5228, respectively. Regarding the ROC analysis of OS (Figure 2B), a tumor diameter ≥6 cm, an AFP level >1,000 ng/mL, donor gender, and a composite index combining the above three predictors were included. The highest AUC of 0.6752 was achieved by the composite index, followed by AFP level (AUC =0.5771), and tumor diameter (AUC =0.5730). The AUC of donor gender was 0.5320. The survival outcomes of patients but not their survival time were included in the ROC analysis, and the sample size was limited, factors which may partly explain the relatively low AUCs for OS. Construction and validation of a risk factor model for predicting DFS. (A) A tumor diameter ≥6 cm, an AFP level >1,000 ng/mL, lack of metformin administration, operative blood loss, donor gender, and a composite index combining the above five predictors were included in the ROC analysis for DFS. A composite index was established using multivariate regression analysis. The composite index achieved the highest AUC of 0.7088. (B) As for the ROC of OS, a tumor diameter ≥6 cm, an AFP level >1,000 ng/mL, donor gender, and a composite index combining the above three predictors were examined. The composite index achieved the highest AUC of 0.6752. (C) A tumor diameter ≥6 cm, an AFP level >1,000 ng/mL, lack of metformin administration, operative blood loss, and donor gender were used to construct a nomogram for predicting DFS, the concordance index of which was 0.708. The actual DFS and the ideal DFS are shown as different lines in the calibration plot. These two lines are closely aligned, demonstrating good calibration for 1- (D) and 5-year DFS (E). AFP, alpha fetoprotein; AUC, area under the curve; DFS, disease-free survival; OS, overall survival; ROC, receiver operating characteristic. As indicated in Table 3, a tumor diameter ≥6 cm, an AFP level >1,000 ng/mL, lack of metformin administration, operative blood loss, and donor gender were selected for the construction of a nomogram predicting DFS, which was evaluated with the concordance index. The concordance index is a measure of the predictive accuracy of the model being tested and ranges from 0.5 (completely random prediction) to 1 (perfect prediction). The concordance index for the nomogram developed in this study was 0.708 (Figure 2C). The actual DFS and the ideal DFS are shown as different lines in the calibration plot in Figure 2D,2E; these two lines were closely aligned, demonstrating good calibration. A tumor diameter ≥6 cm, an AFP level >1,000 ng/mL, and female donor gender were included in the nomogram for predicting OS (Figure 3A). Its concordance index was 0.637, with moderate calibration in 1-year survival (Figure 3B) and 5-year survival (Figure 3C). Construction and validation of the risk factor model for predicting OS. (A) In accordance with the results of multivariate analysis for OS, a tumor diameter ≥6 cm, an AFP level >1,000 ng/mL, and donor gender were included in a nomogram for predicting OS, the concordance index of which was 0.637. The actual OS and the ideal OS are shown as different lines in the calibration plot. These two lines are closely aligned, demonstrating good calibration for 1- (B) and 5-year OS (C). AFP, alpha fetoprotein; OS, overall survival. Discussion In this study, we were able to identify predictors of long-term prognosis of patients with HCC exceeding the Hangzhou criteria undergoing LT. We subsequently constructed predictive models to evaluate the risk of patients selected for LT. Our results could aid in the selection of patients with HCC who may benefit from LT. As a whole, the cumulative survival proportion of our study population was lower than that of patients receiving LT within the Hangzhou criteria. This is likely attributable to the higher tumor pathological stage, greater microvascular invasion, and stronger tumor biological activity of the patients in our study. HCC is one of the critical indications for LT in China. Reducing recurrence rates following LT for liver cancer and improving the prognosis of these patients remains a key clinical challenge. The selection of appropriate LT recipients is the key to improving the effect of LT for liver cancer (16). The widely used liver transplant recipient criteria set include the Milan criteria and University of California San Francisco (UCSF) criteria, which use the number of tumors and tumor size as the basis for screening recipients. However, these criteria are based mainly on a pretransplant radiological imaging, which does not take tumor biological behavior into consideration and fails to characterize small tumors (17,18). A previous study reported a 30% risk of tumor grade underestimation based on preoperative CT or MRI for individuals with HCC (17). In order to compensate for this deficiency, several biological markers have been proposed to be used in combination with the currently used criteria, for example, albumin messenger RNA (mRNA) (18), osteopontin (19), matrix metalloproteinase-9 (20), and importantly, AFP (21). The Hangzhou criteria were the first to incorporate the AFP level and preoperative histological grade as the criteria for recipient screening, which significantly expands the range of indications for LT without affecting the OS rate and DFS rate; however, there is still a lack of high-level medical evidence supporting this approach (15,22). The Hangzhou criteria allow for some patients outside the Milan criteria to receive LT, with up to 37.5% more patients being eligible to receive LT (23). In China, a large country with a high incidence of liver cancer, there is also a large proportion of patients with advanced liver cancer exceeding the Hangzhou criteria. Early study has shown that patients with advanced liver cancer also have a long tumor-free survival time after LT (22). Properly expanding the selection criteria of patients with liver cancer for LT may improve the prognosis of patients with primary liver cancer in China. One of the factors influencing the prognosis of patients with advanced HCC after surgery is carcinoma recurrence after surgery. Although LT has been widely recognized as a radical treatment for patients with HCC without extrahepatic metastasis, there is sparse evidence reported for patients with advanced HCC due to the exceedingly high recurrence rate. In this study, we suggest that appropriately evaluating the biological activity of tumor may effectively resolve this dilemma. The results of this study showed that AFP level and tumor diameter were both independent risk factors associated with both DFS and OS. AFP level is the most commonly used biomarker in the prognostication of patients with liver cancer (24). The Hangzhou criteria consider an AFP level higher than 400 ng/mL to be an important index for evaluating tumor biological activity (25). In our study, after patients were stratified into two groups according to the preoperative AFP, there were significant differences in both the DFS and OS between the two groups, indicating that AFP is also a critical index for the prognosis of patients exceeding Hangzhou criteria who have undergone LT. Importantly, the 5-year DFS and OS of patients with an AFP level ≤1,000 ng/mL was 29.4% and 22.6%, respectively. Results of long-term survival of these patients were compatible with those undergoing surgical resection, with a 30.5% DFS rate and a 24.0% OS rate according to the investigation of patients with advanced HCC (26). Patients involved in this study exhibited poor postoperative prognosis. By the last follow-up, the tumor recurrence rate of these patients was as high as 94%, while the 5-year survival rate was only 17%. These were mainly associated several factors: In terms of tumor characteristics, the enrolled patients had an average tumor diameter of 7.77 cm and an average number of tumor lesions of 5.49. Although portal vein tumor thrombus was excluded by preoperative imaging examination, postoperative pathology still confirmed the presence of microvascular invasion in a large number of patients. Such high tumor burden and occult vascular invasion significantly increased the risk of postoperative recurrence. In terms of comorbidities and treatment management, 83.9% of the patients were complicated with hepatitis B, and insufficient adherence to anti-hepatitis B treatment before and after surgery further aggravated the risk of virus-related tumor recurrence. Meanwhile, 85.7% of the patients had diabetes mellitus; the state of hyperglycemia and insulin resistance affected the prognosis further by disrupting the hepatic microenvironment and impairing the body’s immune surveillance function. Our results indicated that recipient age >50 years old (P=0.04 in univariate analysis) and operative blood loss >2,000 mL (P=0.01 in multivariate analysis) were risk factor predictive of OS after surgery. The cutoff value of age was 50 years old, meaning that older adult patients might be at greater anesthetic risk and liver injury risk. Higher operative blood loss could worsen the OS of patients, and this might be attributable to the ischemia-reperfusion injury during operation. The impact upon the immune microenvironment due to ischemia–reperfusion injury is likely to persistently interfere with the administration and effectiveness of immunosuppressors (27). Moreover, despite the low proportion of female liver donors, female donor gender was found to be a significant risk factor for a worse DFS (P=0.004) and OS (P<0.001) for patients undergoing LT. However, it is important to note that this observed association may not be directly attributed to donor gender itself. Instead, it could potentially reflect unmeasured donor-recipient compatibility factors that were not captured in our current study, such as human leukocyte antigen matching, donor-recipient age or weight discrepancy, and subtle differences in ABO subtype compatibility. Additionally, the limited sample size of our cohort should be considered when interpreting this finding, as it may also contribute to the observed correlation. Metformin administration was also demonstrated to be protective against tumor recurrence (P=0.008). Metformin is typically administered for patients with type 2 diabetes, and has been widely proven to be directly or indirectly related to the biological activity of different types of cancer including HCC (28,29). However, few investigations have examined the connection between metformin administration and the prognosis of patients with HCC undergoing LT. This antitumor function of metformin might be explained by the remission of diabetes and hepatic steatosis and the modulation of inflammation (30,31). Wabitsch et al. showed metformin reverses NASH-induced CD8+ T cell metabolic exhaustion by upregulating mitochondrial genes, restoring T cell motility and anti-tumor activity—critical for countering immunosuppression-related T cell dysfunction post-LT (32). Mao et al. demonstrated metformin disrupts HCC cell polyamine synthesis via ASS1 downregulation and activates AMPK to reduce steatosis and suppress mTOR, directly inhibiting residual tumor cell proliferation post-LT (33). Feng et al. identified the DOCK1-RAC1 axis as a metformin resistance mediator: metformin promotes DOCK1 phosphorylation to activate RAC1 survival signaling, with low DOCK1 expression predicting metformin benefit, suggesting DOCK1 as a potential biomarker for stratifying LT patients (34). Collectively, these studies link metformins anti-recurrence effect to immune restoration, metabolic reprogramming, and context-dependent sensitivity, supporting its use as an adjuvant therapy for high-risk LT recipients. An analysis of the connection between metformin administration and tumor biology could not be conducted in depth in our study because of the limited sample size, and more detailed, well-designed investigations regarding the mechanism underlying the link between metformin diabetes, HCC, and LT are warranted. This study has several limitations. First, as a single-centre retrospective study, the sample reflects the specific patient population and clinical practices of our centre, limiting generalizability to other institutions with different patient demographics or protocols. Second, retrospective data collection introduced inherent biases: missing data on some confounding factors and inability to control for temporal changes in clinical management may have influenced associations between variables like donor gender and survival. Third, the lack of external validation further constrains the reliability of our findings on metformin’s protective effect against HCC recurrence, as we could not confirm these results in an independent cohort. Conclusions In conclusion, for patients exceeding the Hangzhou criteria, those with an AFP level ≤1,000 ng/mL can achieve better long-term prognosis after LT for HCC. Administration of metformin demonstrated a strong positive correlation with better recurrence-free survival after LT for HCC. The data suggest that patient selection for LT can potentially be expanded, but further multicenter randomized studies are needed to validate our results. Supplementary The article’s supplementary files as Available at https://jgo.amegroups.com/article/view/10.21037/jgo-2025-aw-864/prf